Strange values of Y+ on a Ahmed body

Frepa · March 8, 2020, 08:35

Hi everyone,

Im a beginner in CFD. For the last assignment in my university i have to create a reliable Ahmed body model, using SST model. This means that the y+ should be around 1 as much as possible. After few iterations on the first cell thickness, the distribution of y+ is pretty much ok, but there is two really specific regions where the y+ goes to 0, or very close to, everytime, and i didn't figured out why.
If someone has a clue of what is happening i would really appreciate it !
Here is what i am talking about : https://www.casimages.com/i/200308014028682006.png.html

Thanks in advance for all the people taking the time to answer me.

arnie333 · March 9, 2020, 08:12

This is due to the low velocity (air flow) at areas of re-circulation....similar phenomenon as on the sloped rear end.

Further, low Y+ values can be seen at the flat front face of the Ahmed body due to the fact that the air flow has practically come to a stop.

Regards

Amiga500 · March 9, 2020, 10:02

As mentioned, it appears to be localised recirculation or BL separation.

Its been years since I've looked at Ahmed so could be very wrong here, but I don't recall there being separated zones ahead of the sloped rear - the whole point of the model was to examine that aspect of turbulence models.

(i) Are you running at a very subsonic air speed
(ii) What is your surface roughness?
(iii) Is the nose definitely fully blended to the main body (no sharp curvature changes)?

Frepa · March 9, 2020, 12:42

Hi,

We are running the simulations with a Reynolds of 30 000 with water (imposed), so with the frontal area of the car as characteristic length the average velocity is around 0.3 m/s so yes, that is a very subsonic air speed.

I honestly have no idea about the surface roughness, but we have imposed a no slip condition on all surfaces of the car (maybe is that a bad idea ?).

Finally we made the geometry with a simple sketch extruded so yes the nose should be fully blended to the main body.

Im wondering if these low y+ are really that dramatic, because i can't see how i can fix it.

Thanks again for your answers.

Amiga500 · March 12, 2020, 07:06

No, no - no-slip is fine.

Looks like your fine as your running in water:

https://www.researchgate.net/publica...Porous_Devices

This is run in water at Re 30000, and there are circulation zones in around the areas you see the Y+ deviations.

Now - obviously that is still just comparing numerical to numerical rather than experimentals - but it gives some confidence in what your doing. But I know Bordeaux have good water tunnels with experienced people running them, so its unlikely they would not have some good internal data to compare against.

karachun · March 12, 2020, 08:40

If you use k-w or k-w SST turbulence models then you don’t have min y+ bound. There are no sudden flow changes at y+ <1. At range 0<y+<5 flow is assumed to be laminar.
https://en.wikipedia.org/wiki/Law_of_the_wall
https://www.youtube.com/watch?v=fJDYtEGMgzs

Frepa · March 15, 2020, 05:49

Hi,

Indeed it seemed to be caused by water and the re circulation areas, which i can not do anything about it.

Thanks to everyone who took the time to answer and help, you guys are awesome !

March 8, 2020, 08:35	Strange values of Y+ on a Ahmed body	#1
Frepa New Member Join Date: Mar 2020 Posts: 4 Rep Power: 6	Hi everyone, Im a beginner in CFD. For the last assignment in my university i have to create a reliable Ahmed body model, using SST model. This means that the y+ should be around 1 as much as possible. After few iterations on the first cell thickness, the distribution of y+ is pretty much ok, but there is two really specific regions where the y+ goes to 0, or very close to, everytime, and i didn't figured out why. If someone has a clue of what is happening i would really appreciate it ! Here is what i am talking about : https://www.casimages.com/i/200308014028682006.png.html Thanks in advance for all the people taking the time to answer me.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Ahmed body simulation	miku11	OpenFOAM Running, Solving & CFD	2	December 18, 2020 08:01
[GAMBIT] Ahmed Body 3D mesh	Saima	ANSYS Meshing & Geometry	19	February 10, 2018 12:59
Ahmed Body	andrea.pasquali	OpenFOAM	10	June 18, 2016 04:16
strange node values @ solid/fluid interface - help	JB	FLUENT	2	November 1, 2008 13:04
ahmed body	hsn	CFX	0	September 15, 2008 07:30

March 9, 2020, 08:12		#2
arnie333 New Member Arnie Join Date: Mar 2017 Posts: 27 Rep Power: 9	This is due to the low velocity (air flow) at areas of re-circulation....similar phenomenon as on the sloped rear end. Further, low Y+ values can be seen at the flat front face of the Ahmed body due to the fact that the air flow has practically come to a stop. Regards

March 9, 2020, 10:02		#3
Amiga500 Member Join Date: Mar 2009 Posts: 36 Rep Power: 17	As mentioned, it appears to be localised recirculation or BL separation. Its been years since I've looked at Ahmed so could be very wrong here, but I don't recall there being separated zones ahead of the sloped rear - the whole point of the model was to examine that aspect of turbulence models. (i) Are you running at a very subsonic air speed (ii) What is your surface roughness? (iii) Is the nose definitely fully blended to the main body (no sharp curvature changes)?

March 9, 2020, 12:42		#4
Frepa New Member Join Date: Mar 2020 Posts: 4 Rep Power: 6	Hi, We are running the simulations with a Reynolds of 30 000 with water (imposed), so with the frontal area of the car as characteristic length the average velocity is around 0.3 m/s so yes, that is a very subsonic air speed. I honestly have no idea about the surface roughness, but we have imposed a no slip condition on all surfaces of the car (maybe is that a bad idea ?). Finally we made the geometry with a simple sketch extruded so yes the nose should be fully blended to the main body. Im wondering if these low y+ are really that dramatic, because i can't see how i can fix it. Thanks again for your answers.

March 12, 2020, 07:06		#5
Amiga500 Member Join Date: Mar 2009 Posts: 36 Rep Power: 17	No, no - no-slip is fine. Looks like your fine as your running in water: https://www.researchgate.net/publica...Porous_Devices This is run in water at Re 30000, and there are circulation zones in around the areas you see the Y+ deviations. Now - obviously that is still just comparing numerical to numerical rather than experimentals - but it gives some confidence in what your doing. But I know Bordeaux have good water tunnels with experienced people running them, so its unlikely they would not have some good internal data to compare against.

March 12, 2020, 08:40		#6
karachun Senior Member Join Date: Nov 2015 Posts: 246 Rep Power: 12	If you use k-w or k-w SST turbulence models then you don’t have min y+ bound. There are no sudden flow changes at y+ <1. At range 0<y+<5 flow is assumed to be laminar. https://en.wikipedia.org/wiki/Law_of_the_wall https://www.youtube.com/watch?v=fJDYtEGMgzs

March 15, 2020, 05:49		#7
Frepa New Member Join Date: Mar 2020 Posts: 4 Rep Power: 6	Hi, Indeed it seemed to be caused by water and the re circulation areas, which i can not do anything about it. Thanks to everyone who took the time to answer and help, you guys are awesome !